- Email: [email protected]
From molecular changes to customised therapy
European Journal of Cancer 38 (2002) 333–338 www.ejconline.com
Review
From molecular changes to customised therapy A. Hemminki*,1 Division of Human Gene Therapy and the Gene Therapy Center, University of Alabama at Birmingham, WTI #602, 1824 6th Ave S., Birmingham, AL 35294-3300, USA Received 1 October 2001; accepted 9 October 2001
Abstract The revolution in molecular methods has allowed the development of approaches whereby cancer-specific changes can be utilised for targeted therapies. Gene therapy strategies include mutation compensation for correction of cancer-associated defects, and molecular chemotherapy for delivering toxic substances locally to tumour cells. Viruses which replicate only in tumour cells represent a powerful novel approach undergoing intensive development, with some exciting clinical results. Cancer vaccines are promising, although the final clinical evidence is still pending. Monoclonal antibodies, in conjunction with chemotherapeutics, are becoming standard therapy. Recently, the first small molecular inhibitors have made clinical breakthroughs. Importantly, the large amount of information available on genes differentially expressed in cancer cells, allows correlation of prognosis and treatment responses to molecular changes. Thus, the future of cancer treatment could be customised treatment based on the molecular properties of the tumour, utilising combinations of novel and conventional agents. # 2002 Elsevier Science Ltd. All rights reserved. Keywords: Tumour suppressor genes; Peutz–Jeghers syndrome; Microsatellite instability; Gene therapy; Adenovirus; Enzyme inhibitors; Cancer vaccines; Immunotherapy; Monoclonal antibodies; Cancer
1. Introduction: cancer as a molecular disease The identification of mutations in tumour suppressor and oncogenes as the cause of cancer immediately suggested rational means for treating the disease. If such changes could be taken advantage of to limit the antitumour effect to cancer cells, damage to normal tissues could be minimised. The improvements seen in recent decades in molecular and genetic methods have given powerful tools to researchers. Consequently, we now have large amounts of data on cancer-associated genes, their expression profiles and mutations seen in cancer cells. The next challenge will be to try to convert the knowledge into therapeutic benefit for patients (Fig. 1). An important realisation is that metastatic cancer arises as a sequence of epigenetic and genetic changes, each of which increases the aggressiveness, viabilility or ability of the clone to escape immune defences. The ‘gatekeeper’ hypothesis suggests that each cell type has a guardian, the mutation of which is required for the initiation of carcinogenesis [1]. If confirmed, this phe* Fax: +1-205-975 8565. E-mail address: [email protected] (A. Hemminki). 1 Dr. Akseli Hemminki received the 2001 Young European Cancer Researcher Award presented by the European Association for Cancer Research (EACR).
nomenon could be useful for intervention. Furthermore, taking advantage of themes common to various cancer types, such as TP53 or Rb/p16 pathway mutations or cyclo-oxygenase 2 overexpression, could result in widely applicable antitumour strategies. Many genes important in tumour development have been identified through research on hereditary cancer syndromes. One such example is LKB1, hereditary mutations of which cause Peutz–Jeghers syndrome [2,3], and somatic mutations are seen in sporadic cancers [4–7]. Molecular diagnosis is not a novel concept. In fact, karyotyping has been used for decades for the classification of leukaemias. Importantly, we now have effective tools such as array and sequencing-based methodologies which allow detailed molecular analysis of each tumour. However, there is much work to be done in correlating molecular profiles with treatment responses and prognosis. 12% of colorectal cancers display a phenomenon termed microsatellite instability or MSI [8,9]. The underlying cause of MSI is the biallelic inactivation of mismatch repair genes [10], which are normally responsible for correcting mistakes that arise during DNA replication. 5-fluorouracil (5-FU) acts by interfering with nucleic acid synthesis and thus its effect could be different on MSI+ tumours. Therefore, we followed
0959-8049/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S0959-8049(01)00368-9
334
A. Hemminki / European Journal of Cancer 38 (2002) 333–338
Fig. 1. From molecular changes to customised therapy. A schematic presentation of how molecular-based cancer treatment could be realised. Detection of mutations and epigenetic changes in tumour suppressor and oncogenes could be used for determining the best treatment combination for each patient individually. In addition, molecular correlates could improve the estimation of prognosis.
colorectal cancer patients, who had received adjuvant 5-FU [11]. The 3-year recurrence-free survival of MSI+ patients was 90%, while it was 43% for MSI patients. Furthermore, we performed a preliminary investigation of Stage IV MSI+ patients and the results suggested a good response to 5-FU. These findings suggest that 5-FU is safe and could be very useful for the treatment of MSI+ colorectal cancer patients [12]. Thus, molecular correlates like MSI, the analysis of which is not labour-intensive or expensive, could have a significant effect on treatment response and prognosis. Chemotherapy and radiation therapy are widely used for treatment of metastatic cases of cancer, but their tumour selectivity is relatively low. In contrast, molecular-based therapeutics are designed to be specific for tumour-associated features. Available approaches include gene therapy, monoclonal antibodies, cancer vaccines and small molecular inhibitors (Table 1).
2. Cancer as a target for gene therapy Initially, gene therapy was developed for the treatment of hereditary and metabolic disorders, but many of the early approaches were characterised by excessive optimism. Gene transfer or therapeutic effects were not demonstrated. As summarised in the landmark Orkin– Motulsky report in 1995, basic research into vector
biology was necessary to improve target cell transduction, which remains the main obstacle in the field to date. Meanwhile, the focus of gene therapy shifted from hereditary to more common acquired disorders, and nearly 80% of patients treated in trials now are cancer patients. In addition, with cancer the aim is to kill the target cells, which is much easier than long-term expression of proteins at therapeutic levels. The main approaches that have been used in cancer gene therapy include mutation compensation, molecular chemotherapy, and replication-competent viruses. The most frequent target for mutation compensation has been TP53, one of the most commonly altered genes in human cancers. A vector can be used to introduce the healthy TP53 gene into cancer cells in order to induce cell cycle arrest and apoptosis. This approach has been widely tested in human trials, with excellent safety data, but the efficacy of this approach has been disappointing [13]. Concurrently, the field has advanced significantly and shortcomings of this strategy have become evident. With a non-replicating virus, it is extremely difficult to transduce every cell in a solid tumour, a requirement for efficacy with mutation compensation strategies. Another widely used approach is molecular chemotherapy or gene-directed enzyme prodrug therapy. The classic example is Herpes Simplex virus type I thymidine kinase (TK), which catalyses the phosphorylation of systemically administrable non-toxic ganciclovir (GCV)
335
A. Hemminki / European Journal of Cancer 38 (2002) 333–338 Table 1 Molecular-based therapies undergoing clinical evaluation Approach
Examples
Target
Gene therapy Mutation compensation
Ad-TP53
p53-mutant cells Reintroduction of p53 causes apoptosis Infected cells Direct and bystander anti-tumor effects by converted prodrug p53/p14ARFReplication kills cells and allows mutant cells increased tumor penetration
NSCLC
Trastuzumab Cetuximab
ErbB2 EGFR
With chemotherapy With radiation or chemotherapy
Rituximab
CD20
Alone or with chemotherapy
Breast cancer SCCHN, CRC, pancreatic cancer Lymphomas, leukaemias
Radioimmunotherapy
Tositumomab-I131
CD20
Antibody targets radiation to tumour Non-Hodgkin’s lymphoma
Cancer vaccines
Activation of immune cells
Dendritic cells
Activate with tumour antigen
Molecular chemotherapy Ad-TK Replicative viruses Monoclonal antibodies
Small molecular inhibitors
dl1520
Features
Clinical success
Glioma SCCHN
Melanoma, breast, gastrointestinal cancer Melanoma, renal cancer, CRC
Modulation of tumour cells Tumour cells
Modified tumour cells elicit immune response
STI571
Blocks ATP-binding site
CML, ALL, GIST
Blocks ATP-binding site
NSCLC
ZD1839
BCR/ABL, c-kit, PDGF EGFR, ErbB2
Ad, adenovirus; NSCLC, non-small cell lung cancer; TK, Herpes simplex virus type I thymidine kinase; SCCHN, squamous cell cancer of the head and neck; CRC, colorectal cancer; CML, chronic myeloid leukaemia; ALL, acute lymphatic leukaemia; GIST, gastrointestinal stromal tumour; EGFR, Epidermal Growth Factor Receptor; PDGF, Platelet-Derived Growth Factor; ATP, adenosine triphosphate.
into GCV-monophosphate, which undergoes further modifications by cellular enzymes into toxic GCV-triphosphate. Furthermore, GCV-triphosphate can escape producer cells via gap junctions and thereby confer a bystander effect, allowing cells not directly affected with the vector to be killed. A limitation of molecular chemotherapy is the promiscuity of the vector systems used. Transduction of normal tissues causes toxicity and thus targeting of the vector and/or expression of the gene is crucial. Fortunately, transcriptional and transductional targeting is becoming more sophisticated, and molecular chemotherapy continues to be an attractive option with clinical potential. One of the most exciting current approaches is replication-competent viruses (Fig. 2). Their development has been based on the need for more effective tumour penetration. In theory, subsequent rounds of replication allows these agents to penetrate deeper into the tumour until all of the cells have been infected [14]. Each infected cell can produce thousands of virions and therefore there is a tremendous potential for amplification of the effect. Obviously, it is crucial to limit the replication of the agent to tumour cells. It has been suggested that some viruses, such as Reovirus, Vesicular stomatitis and the Newcastle disease virus, could have intrinsic selectivity for replication in tumours [15–17]. The basis of the tumour selectivity of Reoviruses may be in part due to activation of the Ras pathway and defective interferon pathways could be involved with some of the other viruses, but the basic
biology of these viruses is largely unknown. The most promising examples of replication-competent viruses involve intensively studied viruses with strong oncolytic potential, such as the adenovirus.
3. Adenoviral cancer gene therapy The adenovirus (Ad) is perhaps the most studied viral gene transfer vector, which facilitates the creation of
Fig. 2. Conditionally-replicating adenoviruses. Tumour cell-specific oncolytic cell killing is achieved by deletion of genes superfluous for replication in tumour cells (diamond shapes). Alternatively, tumouror tissue-specific promoters can be used for controlling expression of crucial adenovirus genes. Subsequent rounds of replication allow effective penetration into the tumour.
336
A. Hemminki / European Journal of Cancer 38 (2002) 333–338
recombinant agents. Advantages of Ad include unparalleled transduction efficacy of dividing and quiescent cells, natural tropism to a wide range of epithelial tissues and easy production of high titres [18]. The ratelimiting factor for infection is binding to the primary receptor, the coxsackie-adenovirus receptor (CAR) [13]. Ad DNA is transported to the nucleus, but does not integrate and thus the risk of mutagenesis is small, and while the short-term expression is undesirable for hereditary disease, it is unproblematic in oncology where the goal is to kill infected cells. Ad is very immunogenic which could be useful for inducing an immune response against the tumour. Unfortunately, it is becoming increasingly clear that expression of CAR on primary human tumour cells, in various tumour types, is highly variable and often low [13]. Thus, it is crucial to analyse primary tumour samples in addition to cell lines [19,20]. Importantly, various approaches have been developed to circumvent the dependence on CAR. For example, bivalent retargeting molecules have been used to mediate interactions between Ad fibre and a tumour-associated receptor [21]. More sophisticated targeting can be achieved by genetically incorporating peptides into the Ad capsid [22], so that all newly produced viral particles display the modified tropism. In addition, advanced generation Ad vectors allow non-invasive imaging of gene expression and localisation [23]. This may help produce important correlative data from early phase trials, allowing estimation of potential efficacy before phase II and III trials. Traditionally, Ads in oncology have been used as replication-deficient vectors for transporting genes into tumour cells. Safety data has been extremely good, but in most cases tumour transduction has been insufficient to achieve clinical responses [13]. However, in a glioma clinical trial, patients were randomised to receive AdTK, packaging cells for retrovirus with TK or Ad-LacZ (a marker gene) [24]. Injections were performed into the operation margins, followed by GCV treatment. The patients injected with Ad-TK survived significantly longer than the other groups, suggesting that clinical responses can be achieved even with a relatively basic approach. It seems likely there will be routine applications for non-replicating viruses, when local treatment can be applied and the tumour burden is small. In addition, the enhancement of infectivity and combination treatments could dramatically improve the results.
4. Conditionally-replicating adenoviruses The replication of conditionally-replicating Ads (CRADs) (Fig. 2) is controlled by specifically engineered deletions or tissue-specific promoters [25]. As Ad is an oncolytic virus, cells allowing replication die in the process. ‘Virotherapy’ with Ad was first explored in
cervical cancer patients in 1956, soon after the description of the agent [26]. Tumour responses were frequently seen, but the lack of cures led to few continuation studies. This approach resurfaced in force when a CRAD called dl1520 was characterised as an anticancer agent [27]. dl1520 does not express E1B55kD protein, which participates in the binding and inactivation of p53 in cells infected with Ad. Thus, it was hypothesised that this virus could only replicate in cells in which TP53 or its counterpart p14ARF was mutated. It remains controversial how precisely this holds true in practice. Furthermore, the lack of E1B55kD reduces the replicative ability of the agent, which may have contributed to the low single-agent efficacy seen in clinical trials [28]. In contrast, dl1520 has been very promising in combination with chemotherapeutics [29]. Objective responses (> 50% reduction in tumour size) were seen in 63% of patients with recurrent head and neck cancer, and 26% had complete responses. An alternative approach for CRAD construction is control of the key genes with tissue- or tumour-specific promoters (TSPs). Promoters that have been utilised in this context include prostate-specific antigen, probasin, kallikrein and osteocalcin for prostate cancer, alphafetoprotein for hepatocarcinoma, DF3/MUC1 and pS2 for breast cancer, and many of these agents are undergoing clinical evaluation. Considering the highly variable CAR expression on human primary tumours, it seems likely that CRAD efficacy could be dramatically increased with the enhancement of infectivity. In preclinical models, it has been demonstrated that the oncolytic effect of CRADs directly correlates with infectivity [21,30]. Thus, a crucial determinant of the clinical usefulness of a CRAD is the balance between replicative potency and infectivity on the one hand and specificity on the other.
5. Other cancer gene therapy vehicles Retroviruses have been used extensively in gene therapy, but the results in randomised clinical trials for cancer have been disappointing due to inherent limitations of the vector [24,31]. High titre production of retroviruses is difficult and only dividing cells can be infected. However, there are strategies such as ex vivo transduction of leucocytes with drug resistance genes, for which retroviruses could be well suited. In addition, targeting of retroviruses and the creation of conditionally-replicating retroviruses could make this classic tool more useful in oncology. Lentiviruses are technically a subclass of retroviruses, but they can also infect non-replicating cells. However, high titre production is not easy and as most vectors are based on HIV, safety concerns need to be fully addressed before clinical applications can be considered.
A. Hemminki / European Journal of Cancer 38 (2002) 333–338
Herpes viruses were among the first replication-competent viruses that were utilised for cancer treatment [32,33]. Pathogenicity is reduced by mutating one of the crucial virulence genes, resulting in replication competence only in cycling cells. Therefore, these agents are not truly tumour selective, but could be useful in the context of compartmental disease such as glioma. Unfortunately, current agents seem to possess relatively low oncolytic potential, and more sophisticated constructs may be necessary for adequate antitumour effects in humans. There is no human disease known to be caused by adeno-associated viruses (AAV), but it can infect and integrate its genome into various types of human cells. Therefore, it presents an attractive candidate for applications where long-term gene expression is desired. In oncology, this could be useful with anti-angiogenesis strategies or other approaches where the purpose is not to kill the infected cells, but to use them to produce antitumour molecules. Polio, vaccinia and alphaviruses have been used as gene delivery platforms, but their role in oncology is currently not clearly established. The concept of using lipids and other non-viral agents is attractive, as synthetic design and manufacture could produce less complex agents, which could have more predictable pharmacodynamics and -kinetics. However, this is also their limitation, as biological gene transfer vehicles (viruses) are the optimised products of evolution and it is hard for humans to produce similar efficiency in a few decades. An interesting field is the development of synthetic viruses, i.e. using components of viruses for improving the attributes of synthetic constructs. With increased understanding of the biological details of viral gene transfer, approaches combining components from different agents, biological or synthetic, could become increasingly popular.
6. Monoclonal antibodies The use of monoclonal antibodies for the treatment of cancer is based on the assumption that tumours express certain moieties to a higher degree than normal cells. These epitopes could then be targeted with antibodies, with the antitumour effect produced by complement- or cell-mediated mechanisms. Alternatively, monoclonal antibodies can be conjugated with toxins, prodrug converting enzymes, or radioactive molecules. Recent clinical successes have validated the approach (Table 1).
337
nised by the immune system. By definition, any growing cancer will have acquired immune tolerance for any heterologous epitopes. Cancer vaccination is based on the assumption that this tolerance can be broken with appropriate stimuli. Current strategies are usually based on one of two approaches. Either cancer cells are modified to secrete a stimulatory molecule (e.g. IL-2) so that they can be recognised by the immune system, or immune cells are modified to recognise tumour epitopes. Typically, the former is performed ex vivo and the latter can be done in vivo by transducing, e.g. cutaneous dendritic cells with tumour epitopes. A number of promising trials have been reported recently [34], and clinical applications could be imminent.
8. Small molecular inhibitors Typically, small molecular inhibitors compete with molecules such as adenosine triphosphate (ATP) for binding sites and reduce activity of tumour-associated receptors or kinases (Table 1). The BCR/ABL fusion protein is the hallmark of chronic myeloid leukaemia, and treatment with STI571 resulted in complete haematological response in 53/54 patients [35]. Expression of c-kit characterises gastrointestinal stromal tumours, and STI571 results in dramatic responses [36].
9. Conclusions The emergence of sophisticated tools for molecular and genetic analyses of cancers has created the concept of molecular-based treatment. A common denominator for the various strategies is detection of tumour-associated changes and subsequent treatment with the appropriate agents, often utilising combinations. Monoclonal antibodies and small molecular inhibitors have displayed clinical benefit and were quickly accepted for the treatment of patients. Gene therapy and cancer vaccines have displayed tremendous promise in trials and the final clinical breakthroughs are imminent. One of the crucial challenges will be realising the full potential of molecular-based agents in routine clinical practice. A requirement for rational molecular-based therapy must be the molecular analysis of tumours and consecutive informed decisions on which agents to use. There is little doubt that current developments are just the tip of the iceberg and thus the future of molecularbased therapies for cancer looks very exciting indeed.
7. Cancer vaccines Acknowledgements Cancers arise from human cells, but the plethora of mutations associated with the malignant phenotype also produces proteins that would be expected to be recog-
This work was supported by the Damon RunyonWalter Winchell Cancer Research Fund, the Sigrid
338
A. Hemminki / European Journal of Cancer 38 (2002) 333–338
Juselius Foundation, the Emil Aaltonen Foundation and the Maud Kuistila Foundation.
References 1. Kinzler KW, Vogelstein B. Cancer-susceptibility genes. Gatekeepers and caretakers. Nature 1997, 386, 761–763. 2. Hemminki A, Tomlinson I, Markie D, et al. Localization of a susceptibility locus for Peutz-Jeghers syndrome to 19p using comparative genomic hybridization and targeted linkage analysis. Nat Genet 1997, 15, 87–90. 3. Hemminki A, Markie D, Tomlinson I, et al. A serine/threonine kinase gene defective in Peutz-Jeghers syndrome. Nature 1998, 391, 184–187. 4. Avizienyte E, Loukola A, Roth S, et al. LKB1 somatic mutations in sporadic tumors. Am J Pathol 1999, 154, 677–681. 5. Avizienyte E, Roth S, Loukola A, et al. Somatic mutations in LKB1 are rare in sporadic colorectal and testicular tumors. Cancer Res 1998, 58, 2087–2090. 6. Hemminki A. The molecular basis and clinical aspects of PeutzJeghers syndrome. Cell Mol Life Sci 1999, 55, 735–750. 7. Ylikorkala A, Avizienyte E, Tomlinson IP, et al. Mutations and impaired function of LKB1 in familial and non-familial PeutzJeghers syndrome and a sporadic testicular cancer. Hum Mol Genet 1999, 8, 45–51. 8. Canzian F, Salovaara R, Hemminki A, et al. Semiautomated assessment of loss of heterozygosity and replication error in tumors. Cancer Res 1996, 56, 3331–3337. 9. Aaltonen LA, Salovaara R, Kristo P, et al. Incidence of hereditary nonpolyposis colorectal cancer and the feasibility of molecular screening for the disease. N Engl J Med 1998, 338, 1481– 1487. 10. Hemminki A, Peltomaki P, Mecklin JP, et al. Loss of the wild type MLH1 gene is a feature of hereditary nonpolyposis colorectal cancer. Nat Genet 1994, 8, 405–410. 11. Hemminki A, Mecklin JP, Jarvinen H, Aaltonen LA, Joensuu H. Microsatellite instability is a favorable prognostic indicator in patients with colorectal cancer receiving chemotherapy. Gastroenterology 2000, 119, 921–928. 12. Hemminki A, Joensuu H. Microsatellite instability as a molecular marker for very good survival in colorectal cancer patients receiving adjuvant chemotherapy. Gastroenterology 2001, 120, 1309–1310. 13. Hemminki A, Alvarez RD. Adenoviruses in oncology. A viable option? BioDrugs 2001, 16. 14. Kirn D, Martuza RL, Zwiebel J. Replication-selective virotherapy for cancer: biological principles, risk management and future directions. Nat Med 2001, 7, 781–787. 15. Coffey MC, Strong JE, Forsyth PA, Lee PW. Reovirus therapy of tumors with activated Ras pathway. Science 1998, 282, 1332– 1334. 16. Balachandran S, Porosnicu M, Barber GN. Oncolytic activity of vesicular stomatitis virus is effective against tumors exhibiting aberrant p53, Ras, or myc function and involves the induction of apoptosis. J Virol 2001, 75, 3474–3479. 17. Nelson NJ. Scientific interest in Newcastle disease virus is reviving. J Natl Cancer Inst 1999, 91, 1708–1710. 18. Russell WC. Update on adenovirus and its vectors. J Gen Virol 2000, 81, 2573–2604.
19. Barker SD, Casado E, Gomez-Navarro J, et al. An immunomagnetic-based method for the purification of ovarian cancer cells from patient-derived ascites. Gynecol Oncol 2001, 82, 57–63. 20. Casado E, Gomez-Navarro J, Yamamoto M, et al. Strategies to accomplish targeted expression of transgenes in ovarian cancer for molecular therapeutic applications. Clin Cancer Res 2001, 7, 2496–2504. 21. Hemminki A, Dmitriev I, Liu B, Desmond RA, Alemany R, Curiel DT. Targeting oncolytic adenoviral agents to the epidermal growth factor pathway with a secretory fusion molecule. Cancer Res 2001, 61, 6377–6381. 22. Mahasreshti PJ, Navarro JG, Kataram M, et al. Adenovirusmediated Soluble FLT-1 Gene Therapy for Ovarian Carcinoma. Clin Cancer Res 2001, 7, 2057–2066. 23. Hemminki A, Belousova N, Zinn KR, et al. An adenovirus with enhanced infectivity mediates molecular chemotherapy of ovarian cancer cells and allows imaging of gene expression. Mol Ther 2001, 4, 223–231. 24. Sandmair AM, Loimas S, Puranen P, et al. Thymidine kinase gene therapy for human malignant glioma, using replicationdeficient retroviruses or adenoviruses. Hum Gene Ther 2000, 11, 2197–2205. 25. Alemany R, Balague C, Curiel DT. Replicative adenoviruses for cancer therapy. Nat Biotechnol 2000, 18, 723–727. 26. Smith RR, Huebner RJ, Rowe WP, Schatten WE, Thomas LB. Studies on the use of viruses in the treatment of carcinoma of the cervix. Cancer 1956, 9, 1211–1218. 27. Bischoff JR, Kirn DH, Williams A, et al. An adenovirus mutant that replicates selectively in p53-deficient human tumor cells. Science 1996, 274, 373–376. 28. Kirn D. Clinical research results with dl1520 (Onyx-015), a replication selective adenovirus for the treatment of cancer: what have we learned? Gene Ther 2001, 8, 89–98. 29. Khuri FR, Nemunaitis J, Ganly I, et al. A controlled trial of intratumoral ONYX-015, a selectively-replicating adenovirus, in combination with cisplatin and 5-fluorouracil in patients with recurrent head and neck cancer. Nature Med 2000, 6, 879–885. 30. Douglas JT, Kim M, Sumerel LA, Carey DE, Curiel DT. Efficient oncolysis by a replicating adenovirus (ad) in vivo is critically dependent on tumor expression of primary ad receptors. Cancer Res 2001, 61, 813–817. 31. Rainov NG, Group GIS. A phase III clinical evaluation of herpes simplex virus type I thymidine kinase and ganciclovir gene therapy as an adjuvant to surgical resection and radiation in adults with previously untreated glioblastoma multiforme. Hum Gene Ther 2000, 11, 2389–2401. 32. Markert JM, Medlock MD, Rabkin SD, et al. Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: results of a phase I trial. Gene Ther 2000, 7, 867–874. 33. Rampling R, Cruickshank G, Papanastassiou V, et al. Toxicity evaluation of replication-competent herpes simplex virus (ICP 34.5 null mutant 1716) in patients with recurrent malignant glioma. Gene Ther 2000, 7, 859–866. 34. Moingeon P. Cancer vaccines. Vaccine 2001, 19, 1305–1326. 35. Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med 2001, 344, 1031–1037. 36. Joensuu H, Roberts PJ, Sarlomo-Rikala M, et al. Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N Engl J Med 2001, 344, 1052–1056.
Review
From molecular changes to customised therapy A. Hemminki*,1 Division of Human Gene Therapy and the Gene Therapy Center, University of Alabama at Birmingham, WTI #602, 1824 6th Ave S., Birmingham, AL 35294-3300, USA Received 1 October 2001; accepted 9 October 2001
Abstract The revolution in molecular methods has allowed the development of approaches whereby cancer-specific changes can be utilised for targeted therapies. Gene therapy strategies include mutation compensation for correction of cancer-associated defects, and molecular chemotherapy for delivering toxic substances locally to tumour cells. Viruses which replicate only in tumour cells represent a powerful novel approach undergoing intensive development, with some exciting clinical results. Cancer vaccines are promising, although the final clinical evidence is still pending. Monoclonal antibodies, in conjunction with chemotherapeutics, are becoming standard therapy. Recently, the first small molecular inhibitors have made clinical breakthroughs. Importantly, the large amount of information available on genes differentially expressed in cancer cells, allows correlation of prognosis and treatment responses to molecular changes. Thus, the future of cancer treatment could be customised treatment based on the molecular properties of the tumour, utilising combinations of novel and conventional agents. # 2002 Elsevier Science Ltd. All rights reserved. Keywords: Tumour suppressor genes; Peutz–Jeghers syndrome; Microsatellite instability; Gene therapy; Adenovirus; Enzyme inhibitors; Cancer vaccines; Immunotherapy; Monoclonal antibodies; Cancer
1. Introduction: cancer as a molecular disease The identification of mutations in tumour suppressor and oncogenes as the cause of cancer immediately suggested rational means for treating the disease. If such changes could be taken advantage of to limit the antitumour effect to cancer cells, damage to normal tissues could be minimised. The improvements seen in recent decades in molecular and genetic methods have given powerful tools to researchers. Consequently, we now have large amounts of data on cancer-associated genes, their expression profiles and mutations seen in cancer cells. The next challenge will be to try to convert the knowledge into therapeutic benefit for patients (Fig. 1). An important realisation is that metastatic cancer arises as a sequence of epigenetic and genetic changes, each of which increases the aggressiveness, viabilility or ability of the clone to escape immune defences. The ‘gatekeeper’ hypothesis suggests that each cell type has a guardian, the mutation of which is required for the initiation of carcinogenesis [1]. If confirmed, this phe* Fax: +1-205-975 8565. E-mail address: [email protected] (A. Hemminki). 1 Dr. Akseli Hemminki received the 2001 Young European Cancer Researcher Award presented by the European Association for Cancer Research (EACR).
nomenon could be useful for intervention. Furthermore, taking advantage of themes common to various cancer types, such as TP53 or Rb/p16 pathway mutations or cyclo-oxygenase 2 overexpression, could result in widely applicable antitumour strategies. Many genes important in tumour development have been identified through research on hereditary cancer syndromes. One such example is LKB1, hereditary mutations of which cause Peutz–Jeghers syndrome [2,3], and somatic mutations are seen in sporadic cancers [4–7]. Molecular diagnosis is not a novel concept. In fact, karyotyping has been used for decades for the classification of leukaemias. Importantly, we now have effective tools such as array and sequencing-based methodologies which allow detailed molecular analysis of each tumour. However, there is much work to be done in correlating molecular profiles with treatment responses and prognosis. 12% of colorectal cancers display a phenomenon termed microsatellite instability or MSI [8,9]. The underlying cause of MSI is the biallelic inactivation of mismatch repair genes [10], which are normally responsible for correcting mistakes that arise during DNA replication. 5-fluorouracil (5-FU) acts by interfering with nucleic acid synthesis and thus its effect could be different on MSI+ tumours. Therefore, we followed
0959-8049/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S0959-8049(01)00368-9
334
A. Hemminki / European Journal of Cancer 38 (2002) 333–338
Fig. 1. From molecular changes to customised therapy. A schematic presentation of how molecular-based cancer treatment could be realised. Detection of mutations and epigenetic changes in tumour suppressor and oncogenes could be used for determining the best treatment combination for each patient individually. In addition, molecular correlates could improve the estimation of prognosis.
colorectal cancer patients, who had received adjuvant 5-FU [11]. The 3-year recurrence-free survival of MSI+ patients was 90%, while it was 43% for MSI patients. Furthermore, we performed a preliminary investigation of Stage IV MSI+ patients and the results suggested a good response to 5-FU. These findings suggest that 5-FU is safe and could be very useful for the treatment of MSI+ colorectal cancer patients [12]. Thus, molecular correlates like MSI, the analysis of which is not labour-intensive or expensive, could have a significant effect on treatment response and prognosis. Chemotherapy and radiation therapy are widely used for treatment of metastatic cases of cancer, but their tumour selectivity is relatively low. In contrast, molecular-based therapeutics are designed to be specific for tumour-associated features. Available approaches include gene therapy, monoclonal antibodies, cancer vaccines and small molecular inhibitors (Table 1).
2. Cancer as a target for gene therapy Initially, gene therapy was developed for the treatment of hereditary and metabolic disorders, but many of the early approaches were characterised by excessive optimism. Gene transfer or therapeutic effects were not demonstrated. As summarised in the landmark Orkin– Motulsky report in 1995, basic research into vector
biology was necessary to improve target cell transduction, which remains the main obstacle in the field to date. Meanwhile, the focus of gene therapy shifted from hereditary to more common acquired disorders, and nearly 80% of patients treated in trials now are cancer patients. In addition, with cancer the aim is to kill the target cells, which is much easier than long-term expression of proteins at therapeutic levels. The main approaches that have been used in cancer gene therapy include mutation compensation, molecular chemotherapy, and replication-competent viruses. The most frequent target for mutation compensation has been TP53, one of the most commonly altered genes in human cancers. A vector can be used to introduce the healthy TP53 gene into cancer cells in order to induce cell cycle arrest and apoptosis. This approach has been widely tested in human trials, with excellent safety data, but the efficacy of this approach has been disappointing [13]. Concurrently, the field has advanced significantly and shortcomings of this strategy have become evident. With a non-replicating virus, it is extremely difficult to transduce every cell in a solid tumour, a requirement for efficacy with mutation compensation strategies. Another widely used approach is molecular chemotherapy or gene-directed enzyme prodrug therapy. The classic example is Herpes Simplex virus type I thymidine kinase (TK), which catalyses the phosphorylation of systemically administrable non-toxic ganciclovir (GCV)
335
A. Hemminki / European Journal of Cancer 38 (2002) 333–338 Table 1 Molecular-based therapies undergoing clinical evaluation Approach
Examples
Target
Gene therapy Mutation compensation
Ad-TP53
p53-mutant cells Reintroduction of p53 causes apoptosis Infected cells Direct and bystander anti-tumor effects by converted prodrug p53/p14ARFReplication kills cells and allows mutant cells increased tumor penetration
NSCLC
Trastuzumab Cetuximab
ErbB2 EGFR
With chemotherapy With radiation or chemotherapy
Rituximab
CD20
Alone or with chemotherapy
Breast cancer SCCHN, CRC, pancreatic cancer Lymphomas, leukaemias
Radioimmunotherapy
Tositumomab-I131
CD20
Antibody targets radiation to tumour Non-Hodgkin’s lymphoma
Cancer vaccines
Activation of immune cells
Dendritic cells
Activate with tumour antigen
Molecular chemotherapy Ad-TK Replicative viruses Monoclonal antibodies
Small molecular inhibitors
dl1520
Features
Clinical success
Glioma SCCHN
Melanoma, breast, gastrointestinal cancer Melanoma, renal cancer, CRC
Modulation of tumour cells Tumour cells
Modified tumour cells elicit immune response
STI571
Blocks ATP-binding site
CML, ALL, GIST
Blocks ATP-binding site
NSCLC
ZD1839
BCR/ABL, c-kit, PDGF EGFR, ErbB2
Ad, adenovirus; NSCLC, non-small cell lung cancer; TK, Herpes simplex virus type I thymidine kinase; SCCHN, squamous cell cancer of the head and neck; CRC, colorectal cancer; CML, chronic myeloid leukaemia; ALL, acute lymphatic leukaemia; GIST, gastrointestinal stromal tumour; EGFR, Epidermal Growth Factor Receptor; PDGF, Platelet-Derived Growth Factor; ATP, adenosine triphosphate.
into GCV-monophosphate, which undergoes further modifications by cellular enzymes into toxic GCV-triphosphate. Furthermore, GCV-triphosphate can escape producer cells via gap junctions and thereby confer a bystander effect, allowing cells not directly affected with the vector to be killed. A limitation of molecular chemotherapy is the promiscuity of the vector systems used. Transduction of normal tissues causes toxicity and thus targeting of the vector and/or expression of the gene is crucial. Fortunately, transcriptional and transductional targeting is becoming more sophisticated, and molecular chemotherapy continues to be an attractive option with clinical potential. One of the most exciting current approaches is replication-competent viruses (Fig. 2). Their development has been based on the need for more effective tumour penetration. In theory, subsequent rounds of replication allows these agents to penetrate deeper into the tumour until all of the cells have been infected [14]. Each infected cell can produce thousands of virions and therefore there is a tremendous potential for amplification of the effect. Obviously, it is crucial to limit the replication of the agent to tumour cells. It has been suggested that some viruses, such as Reovirus, Vesicular stomatitis and the Newcastle disease virus, could have intrinsic selectivity for replication in tumours [15–17]. The basis of the tumour selectivity of Reoviruses may be in part due to activation of the Ras pathway and defective interferon pathways could be involved with some of the other viruses, but the basic
biology of these viruses is largely unknown. The most promising examples of replication-competent viruses involve intensively studied viruses with strong oncolytic potential, such as the adenovirus.
3. Adenoviral cancer gene therapy The adenovirus (Ad) is perhaps the most studied viral gene transfer vector, which facilitates the creation of
Fig. 2. Conditionally-replicating adenoviruses. Tumour cell-specific oncolytic cell killing is achieved by deletion of genes superfluous for replication in tumour cells (diamond shapes). Alternatively, tumouror tissue-specific promoters can be used for controlling expression of crucial adenovirus genes. Subsequent rounds of replication allow effective penetration into the tumour.
336
A. Hemminki / European Journal of Cancer 38 (2002) 333–338
recombinant agents. Advantages of Ad include unparalleled transduction efficacy of dividing and quiescent cells, natural tropism to a wide range of epithelial tissues and easy production of high titres [18]. The ratelimiting factor for infection is binding to the primary receptor, the coxsackie-adenovirus receptor (CAR) [13]. Ad DNA is transported to the nucleus, but does not integrate and thus the risk of mutagenesis is small, and while the short-term expression is undesirable for hereditary disease, it is unproblematic in oncology where the goal is to kill infected cells. Ad is very immunogenic which could be useful for inducing an immune response against the tumour. Unfortunately, it is becoming increasingly clear that expression of CAR on primary human tumour cells, in various tumour types, is highly variable and often low [13]. Thus, it is crucial to analyse primary tumour samples in addition to cell lines [19,20]. Importantly, various approaches have been developed to circumvent the dependence on CAR. For example, bivalent retargeting molecules have been used to mediate interactions between Ad fibre and a tumour-associated receptor [21]. More sophisticated targeting can be achieved by genetically incorporating peptides into the Ad capsid [22], so that all newly produced viral particles display the modified tropism. In addition, advanced generation Ad vectors allow non-invasive imaging of gene expression and localisation [23]. This may help produce important correlative data from early phase trials, allowing estimation of potential efficacy before phase II and III trials. Traditionally, Ads in oncology have been used as replication-deficient vectors for transporting genes into tumour cells. Safety data has been extremely good, but in most cases tumour transduction has been insufficient to achieve clinical responses [13]. However, in a glioma clinical trial, patients were randomised to receive AdTK, packaging cells for retrovirus with TK or Ad-LacZ (a marker gene) [24]. Injections were performed into the operation margins, followed by GCV treatment. The patients injected with Ad-TK survived significantly longer than the other groups, suggesting that clinical responses can be achieved even with a relatively basic approach. It seems likely there will be routine applications for non-replicating viruses, when local treatment can be applied and the tumour burden is small. In addition, the enhancement of infectivity and combination treatments could dramatically improve the results.
4. Conditionally-replicating adenoviruses The replication of conditionally-replicating Ads (CRADs) (Fig. 2) is controlled by specifically engineered deletions or tissue-specific promoters [25]. As Ad is an oncolytic virus, cells allowing replication die in the process. ‘Virotherapy’ with Ad was first explored in
cervical cancer patients in 1956, soon after the description of the agent [26]. Tumour responses were frequently seen, but the lack of cures led to few continuation studies. This approach resurfaced in force when a CRAD called dl1520 was characterised as an anticancer agent [27]. dl1520 does not express E1B55kD protein, which participates in the binding and inactivation of p53 in cells infected with Ad. Thus, it was hypothesised that this virus could only replicate in cells in which TP53 or its counterpart p14ARF was mutated. It remains controversial how precisely this holds true in practice. Furthermore, the lack of E1B55kD reduces the replicative ability of the agent, which may have contributed to the low single-agent efficacy seen in clinical trials [28]. In contrast, dl1520 has been very promising in combination with chemotherapeutics [29]. Objective responses (> 50% reduction in tumour size) were seen in 63% of patients with recurrent head and neck cancer, and 26% had complete responses. An alternative approach for CRAD construction is control of the key genes with tissue- or tumour-specific promoters (TSPs). Promoters that have been utilised in this context include prostate-specific antigen, probasin, kallikrein and osteocalcin for prostate cancer, alphafetoprotein for hepatocarcinoma, DF3/MUC1 and pS2 for breast cancer, and many of these agents are undergoing clinical evaluation. Considering the highly variable CAR expression on human primary tumours, it seems likely that CRAD efficacy could be dramatically increased with the enhancement of infectivity. In preclinical models, it has been demonstrated that the oncolytic effect of CRADs directly correlates with infectivity [21,30]. Thus, a crucial determinant of the clinical usefulness of a CRAD is the balance between replicative potency and infectivity on the one hand and specificity on the other.
5. Other cancer gene therapy vehicles Retroviruses have been used extensively in gene therapy, but the results in randomised clinical trials for cancer have been disappointing due to inherent limitations of the vector [24,31]. High titre production of retroviruses is difficult and only dividing cells can be infected. However, there are strategies such as ex vivo transduction of leucocytes with drug resistance genes, for which retroviruses could be well suited. In addition, targeting of retroviruses and the creation of conditionally-replicating retroviruses could make this classic tool more useful in oncology. Lentiviruses are technically a subclass of retroviruses, but they can also infect non-replicating cells. However, high titre production is not easy and as most vectors are based on HIV, safety concerns need to be fully addressed before clinical applications can be considered.
A. Hemminki / European Journal of Cancer 38 (2002) 333–338
Herpes viruses were among the first replication-competent viruses that were utilised for cancer treatment [32,33]. Pathogenicity is reduced by mutating one of the crucial virulence genes, resulting in replication competence only in cycling cells. Therefore, these agents are not truly tumour selective, but could be useful in the context of compartmental disease such as glioma. Unfortunately, current agents seem to possess relatively low oncolytic potential, and more sophisticated constructs may be necessary for adequate antitumour effects in humans. There is no human disease known to be caused by adeno-associated viruses (AAV), but it can infect and integrate its genome into various types of human cells. Therefore, it presents an attractive candidate for applications where long-term gene expression is desired. In oncology, this could be useful with anti-angiogenesis strategies or other approaches where the purpose is not to kill the infected cells, but to use them to produce antitumour molecules. Polio, vaccinia and alphaviruses have been used as gene delivery platforms, but their role in oncology is currently not clearly established. The concept of using lipids and other non-viral agents is attractive, as synthetic design and manufacture could produce less complex agents, which could have more predictable pharmacodynamics and -kinetics. However, this is also their limitation, as biological gene transfer vehicles (viruses) are the optimised products of evolution and it is hard for humans to produce similar efficiency in a few decades. An interesting field is the development of synthetic viruses, i.e. using components of viruses for improving the attributes of synthetic constructs. With increased understanding of the biological details of viral gene transfer, approaches combining components from different agents, biological or synthetic, could become increasingly popular.
6. Monoclonal antibodies The use of monoclonal antibodies for the treatment of cancer is based on the assumption that tumours express certain moieties to a higher degree than normal cells. These epitopes could then be targeted with antibodies, with the antitumour effect produced by complement- or cell-mediated mechanisms. Alternatively, monoclonal antibodies can be conjugated with toxins, prodrug converting enzymes, or radioactive molecules. Recent clinical successes have validated the approach (Table 1).
337
nised by the immune system. By definition, any growing cancer will have acquired immune tolerance for any heterologous epitopes. Cancer vaccination is based on the assumption that this tolerance can be broken with appropriate stimuli. Current strategies are usually based on one of two approaches. Either cancer cells are modified to secrete a stimulatory molecule (e.g. IL-2) so that they can be recognised by the immune system, or immune cells are modified to recognise tumour epitopes. Typically, the former is performed ex vivo and the latter can be done in vivo by transducing, e.g. cutaneous dendritic cells with tumour epitopes. A number of promising trials have been reported recently [34], and clinical applications could be imminent.
8. Small molecular inhibitors Typically, small molecular inhibitors compete with molecules such as adenosine triphosphate (ATP) for binding sites and reduce activity of tumour-associated receptors or kinases (Table 1). The BCR/ABL fusion protein is the hallmark of chronic myeloid leukaemia, and treatment with STI571 resulted in complete haematological response in 53/54 patients [35]. Expression of c-kit characterises gastrointestinal stromal tumours, and STI571 results in dramatic responses [36].
9. Conclusions The emergence of sophisticated tools for molecular and genetic analyses of cancers has created the concept of molecular-based treatment. A common denominator for the various strategies is detection of tumour-associated changes and subsequent treatment with the appropriate agents, often utilising combinations. Monoclonal antibodies and small molecular inhibitors have displayed clinical benefit and were quickly accepted for the treatment of patients. Gene therapy and cancer vaccines have displayed tremendous promise in trials and the final clinical breakthroughs are imminent. One of the crucial challenges will be realising the full potential of molecular-based agents in routine clinical practice. A requirement for rational molecular-based therapy must be the molecular analysis of tumours and consecutive informed decisions on which agents to use. There is little doubt that current developments are just the tip of the iceberg and thus the future of molecularbased therapies for cancer looks very exciting indeed.
7. Cancer vaccines Acknowledgements Cancers arise from human cells, but the plethora of mutations associated with the malignant phenotype also produces proteins that would be expected to be recog-
This work was supported by the Damon RunyonWalter Winchell Cancer Research Fund, the Sigrid
338
A. Hemminki / European Journal of Cancer 38 (2002) 333–338
Juselius Foundation, the Emil Aaltonen Foundation and the Maud Kuistila Foundation.
References 1. Kinzler KW, Vogelstein B. Cancer-susceptibility genes. Gatekeepers and caretakers. Nature 1997, 386, 761–763. 2. Hemminki A, Tomlinson I, Markie D, et al. Localization of a susceptibility locus for Peutz-Jeghers syndrome to 19p using comparative genomic hybridization and targeted linkage analysis. Nat Genet 1997, 15, 87–90. 3. Hemminki A, Markie D, Tomlinson I, et al. A serine/threonine kinase gene defective in Peutz-Jeghers syndrome. Nature 1998, 391, 184–187. 4. Avizienyte E, Loukola A, Roth S, et al. LKB1 somatic mutations in sporadic tumors. Am J Pathol 1999, 154, 677–681. 5. Avizienyte E, Roth S, Loukola A, et al. Somatic mutations in LKB1 are rare in sporadic colorectal and testicular tumors. Cancer Res 1998, 58, 2087–2090. 6. Hemminki A. The molecular basis and clinical aspects of PeutzJeghers syndrome. Cell Mol Life Sci 1999, 55, 735–750. 7. Ylikorkala A, Avizienyte E, Tomlinson IP, et al. Mutations and impaired function of LKB1 in familial and non-familial PeutzJeghers syndrome and a sporadic testicular cancer. Hum Mol Genet 1999, 8, 45–51. 8. Canzian F, Salovaara R, Hemminki A, et al. Semiautomated assessment of loss of heterozygosity and replication error in tumors. Cancer Res 1996, 56, 3331–3337. 9. Aaltonen LA, Salovaara R, Kristo P, et al. Incidence of hereditary nonpolyposis colorectal cancer and the feasibility of molecular screening for the disease. N Engl J Med 1998, 338, 1481– 1487. 10. Hemminki A, Peltomaki P, Mecklin JP, et al. Loss of the wild type MLH1 gene is a feature of hereditary nonpolyposis colorectal cancer. Nat Genet 1994, 8, 405–410. 11. Hemminki A, Mecklin JP, Jarvinen H, Aaltonen LA, Joensuu H. Microsatellite instability is a favorable prognostic indicator in patients with colorectal cancer receiving chemotherapy. Gastroenterology 2000, 119, 921–928. 12. Hemminki A, Joensuu H. Microsatellite instability as a molecular marker for very good survival in colorectal cancer patients receiving adjuvant chemotherapy. Gastroenterology 2001, 120, 1309–1310. 13. Hemminki A, Alvarez RD. Adenoviruses in oncology. A viable option? BioDrugs 2001, 16. 14. Kirn D, Martuza RL, Zwiebel J. Replication-selective virotherapy for cancer: biological principles, risk management and future directions. Nat Med 2001, 7, 781–787. 15. Coffey MC, Strong JE, Forsyth PA, Lee PW. Reovirus therapy of tumors with activated Ras pathway. Science 1998, 282, 1332– 1334. 16. Balachandran S, Porosnicu M, Barber GN. Oncolytic activity of vesicular stomatitis virus is effective against tumors exhibiting aberrant p53, Ras, or myc function and involves the induction of apoptosis. J Virol 2001, 75, 3474–3479. 17. Nelson NJ. Scientific interest in Newcastle disease virus is reviving. J Natl Cancer Inst 1999, 91, 1708–1710. 18. Russell WC. Update on adenovirus and its vectors. J Gen Virol 2000, 81, 2573–2604.
19. Barker SD, Casado E, Gomez-Navarro J, et al. An immunomagnetic-based method for the purification of ovarian cancer cells from patient-derived ascites. Gynecol Oncol 2001, 82, 57–63. 20. Casado E, Gomez-Navarro J, Yamamoto M, et al. Strategies to accomplish targeted expression of transgenes in ovarian cancer for molecular therapeutic applications. Clin Cancer Res 2001, 7, 2496–2504. 21. Hemminki A, Dmitriev I, Liu B, Desmond RA, Alemany R, Curiel DT. Targeting oncolytic adenoviral agents to the epidermal growth factor pathway with a secretory fusion molecule. Cancer Res 2001, 61, 6377–6381. 22. Mahasreshti PJ, Navarro JG, Kataram M, et al. Adenovirusmediated Soluble FLT-1 Gene Therapy for Ovarian Carcinoma. Clin Cancer Res 2001, 7, 2057–2066. 23. Hemminki A, Belousova N, Zinn KR, et al. An adenovirus with enhanced infectivity mediates molecular chemotherapy of ovarian cancer cells and allows imaging of gene expression. Mol Ther 2001, 4, 223–231. 24. Sandmair AM, Loimas S, Puranen P, et al. Thymidine kinase gene therapy for human malignant glioma, using replicationdeficient retroviruses or adenoviruses. Hum Gene Ther 2000, 11, 2197–2205. 25. Alemany R, Balague C, Curiel DT. Replicative adenoviruses for cancer therapy. Nat Biotechnol 2000, 18, 723–727. 26. Smith RR, Huebner RJ, Rowe WP, Schatten WE, Thomas LB. Studies on the use of viruses in the treatment of carcinoma of the cervix. Cancer 1956, 9, 1211–1218. 27. Bischoff JR, Kirn DH, Williams A, et al. An adenovirus mutant that replicates selectively in p53-deficient human tumor cells. Science 1996, 274, 373–376. 28. Kirn D. Clinical research results with dl1520 (Onyx-015), a replication selective adenovirus for the treatment of cancer: what have we learned? Gene Ther 2001, 8, 89–98. 29. Khuri FR, Nemunaitis J, Ganly I, et al. A controlled trial of intratumoral ONYX-015, a selectively-replicating adenovirus, in combination with cisplatin and 5-fluorouracil in patients with recurrent head and neck cancer. Nature Med 2000, 6, 879–885. 30. Douglas JT, Kim M, Sumerel LA, Carey DE, Curiel DT. Efficient oncolysis by a replicating adenovirus (ad) in vivo is critically dependent on tumor expression of primary ad receptors. Cancer Res 2001, 61, 813–817. 31. Rainov NG, Group GIS. A phase III clinical evaluation of herpes simplex virus type I thymidine kinase and ganciclovir gene therapy as an adjuvant to surgical resection and radiation in adults with previously untreated glioblastoma multiforme. Hum Gene Ther 2000, 11, 2389–2401. 32. Markert JM, Medlock MD, Rabkin SD, et al. Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: results of a phase I trial. Gene Ther 2000, 7, 867–874. 33. Rampling R, Cruickshank G, Papanastassiou V, et al. Toxicity evaluation of replication-competent herpes simplex virus (ICP 34.5 null mutant 1716) in patients with recurrent malignant glioma. Gene Ther 2000, 7, 859–866. 34. Moingeon P. Cancer vaccines. Vaccine 2001, 19, 1305–1326. 35. Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med 2001, 344, 1031–1037. 36. Joensuu H, Roberts PJ, Sarlomo-Rikala M, et al. Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N Engl J Med 2001, 344, 1052–1056.